Investment Decisions and Falling Cost of Data Analytics

Abstract

Central and local governments around the world are seeking investments from private firms to create smart city solutions. Motivated by this, we model an investor and a local government with a Stackelberg game. The investor has a CARA (or exponential) utility function and she has an option to acquire data dynamically and use that in a Bayesian predictive model to forecast the quality sensitivity of demand before her investment decision. The government chooses the highest price such that demand quantity is above a minimum threshold. We prove comparative statics under the dynamic information acquisition and the Stackelberg game. We show that borrowing-constrained or highly risk-averse investors have low demand for information. Fur- thermore, the falling marginal cost of information driven by smart cities’ datasets and analytics raises investment amounts and leverage, which leads to higher losses during crises. Finally, we show that leverage constraints are effective in mitigating losses during crises.